Complex AI Challenges: Minimizing Admissions Bias, Keeping Degrees Relevant

While University of Pennsylvania researchers, in a far-reaching study, found they could use artificial intelligence (AI) to assess application essays with a minimum of bias, a survey by a leading recruiter predicts that the technology could devalue the four-year degree. The Penn study found AI can be used to assess admissions essays—and probe them for indications of seven personal qualities—with minimal bias. The project, which tested more than 300,000 U.S. student application essays, also achieved “overall accuracy in predicting eventual graduation success,” according to Times Higher Education. Meanwhile, Axios reported on the Talent Connect Summit in October in New York City, attended by 2,000 of the nation’s top recruiters. Executives from LinkedIn, used by 90 percent of recruiters to find job candidates, said: “In AI-driven workplaces, employers will need to treat ‘up-skilling’ [giving current employees additional skills] as a ‘critical priority’ rather than a perk.”‍In LinkedIn’s survey, 72 percent of American executives said they value soft skills (like critical thinking/problem solving, attention to detail, communication, leadership, teamwork, grit) more highly than AI skills. LinkedIn CEO Ryan Roslansky told attendees that as AI automates menial work and some knowledge work, companies will place more focus on “human and people-oriented skills.” That, he said, will “make it virtually impossible for a one-off moment of learning [like a degree] to last an entire career.”‍